High Lead Exposures Resulting from Pottery Production in a Village in Michoacaân State, Mexico

http://www.stockton-press.co.uk

High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico

ROBIN HIBBERT,a ZHIPENG BAI,b JAIME NAVIA,c DANIEL M. KAMMENa,d AND JUNFENG (JIM) ZHANGb a Science, Technology and Environmental Policy (STEP) Program, Woodrow Wilson School of Public and International Affairs, Princeton University, Princeton, New Jersey 08544-1013 b Environmental and Occupational Health Sciences Institute, UMDNJ-Robert Wood Johnson Medical School and Rutgers University, 170 Frelinghuysen Road, Piscataway, New Jersey 08854 c Grupo Interdisciplinario de TecnologIÁa Rural Apropiada (GIRA), AC, Apartado Postal 158, PaÂtzcuaro, MichoacaÂn 61609, Mexico d Energy and Resources Group (ERG), University of California, Berkeley, Berkeley, California 94720-3050

This paper reports findings from a screening study conducted to examine potential lead (Pb) exposures in residents of a Mexican village where Pb oxide continues to be used in ceramic pottery production. Extremely high Pb concentrations were measured in personal and indoor air samples, household surface dust samples, and household soil samples. Personal air Pb concentrations for workers performing pottery firing and glazing were up to 454 g/m3. Results from indoor air samples indicate that airborne Pb concentrations were lower during nonglazing period compared to the glazing period. Soil Pb concentrations measured in 17 homes ranged from 0.39 to 19.8 mg/g. Dust Pb loading on surfaces of household items, hands, and clothes of a worker ranged from 172 to 33 060 g/f t 2. Pb content as high as 2.4 g/g was found in a bean stew cooked in a pot made in the village. Based on these Pb concentrations measured in multiple media and data adapted for exposure contact rates, we have made rough estimates of Pb exposures via inhalation, soil/dust ingestion, and food ingestion. Estimated total daily Pb intake, on average, is 4.0 mg for adults and 3.4 mg for children living in the village. In the total daily intake, a greatest fraction may be contributed by food ingestion and another significant fraction may come from soil/dust ingestion for the children. Although the sample size is small, these measurements indicate a very significant public health problem for the village residents and a large number of other similar communitiesin Mexico. (It was estimated that there are approximately 1.5 million glaze potters.) The Pb exposure is implicated in a number of pervasive health problems in the region, and is the cause for national and international attention. Several recommended solutions to this problem range from personal protection and behavioral changes to introduction of alternative glazes.

Keywords: ceramic glaze, Mexico, Pb exposure, Pb poisoning, pottery.

Introduction the US. This unfortunate fact resulted in the Centers for Disease Control and Prevention (CDC) lowering the acceptable blood lead (BPb) level three times in the last Lead (Pb) is a well-known toxic substance. Due to its 20 years (CDC, 1991). The current action level for BPb is toxicity and environmental prevalence, Pb poisoning has 10 g/dl. Since principal Pb sources such as leaded gasoline been one of the worst environmental threats to children in and Pb-based paints have been removed in the US, a dramatic decrease in BPb levels was reported from 12.8 to 2.8 g/dl between 1976±1980 and 1988±1991 (Harrison 1. Abbreviations: AIHC, American Industrial Health Council; ASTM, American Society for Testing and Materials; BPb, blood lead; CDC, and Laxen, 1981; Pirkle et al., 1994). Most significantly, the Centers for Disease Control and Prevention; DHHS, US Department of percentage of US children aged 1±5 years with BPb levels Health and Human Services; EOHSI, Environmental and Occupational 10 g/dl or greater dropped from 88.2% to 8.9%. This may Health Sciences Institute; EPA, US Environmental Protection Agency; look like a signal for declaring victory in the war against Pb FAA, flame atomic absorption; HUD, US Department of Housing and poisoning. One might hesitate to do so, however, by Urban Development; ICP-MS, inductively coupled plasma-mass spectro- metry; ISO, International Standard Organization; NIST, National Institute examining Pb exposures of some specific populations in of Standards and Technology; Pb, lead. some parts of the world (Diaz-Barriga et al., 1997; 2. Address all correspondence to: Junfeng (Jim) Zhang, Ph.D., Exposure Murgueytio et al., 1998). Measurement and Assessment Division, Environmental and Occupa- A recent study (Fernandez et al., 1997) showed that tional Health Sciences Institute, UMDNJ-Robert Wood Johnson Medical measured BPb levels of ceramic workers in three villages in School and Rutgers University, 170 Frelinghuysen Road, Piscat- away, NJ 08854. Tel.: (732)445-0158. Fax: (732)445-0116. E-mail: MichoacaÂn State of Mexico significantly exceeded US [email protected] recommended safety level of 10 g/dl. The mean BPb Revised 14 December 1998; accepted 23 December 1998. values for 310 female and 146 male workers were 22.9 and Hibbert et al. High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico

23.9 g/dl, respectively. The study, however, included area, which was also an area where children frequently blood samples from residents who did not work with pottery played on the soil floors. A stainless steel spoon was used to in their homes and those who used closed gas ovens in scrape the sample from the ground and then the sample was determining community and age/gender-stratified averages. placed in a Ziploc bag. The soil samples were dried for Therefore, BPb levels for pottery workers might be even analysis at 438C in an oven for 1 week. Portions between 0.2 higher than the community averages reported by Fernandez and 0.8 g were weighed and placed into test tubes (Oak et al. Ridge Centrifuge Tube, 30 ml). The soil samples were then Although Fernandez et al. (1997) found high BPb levels, digested in a microwave oven using a total Pb digestion no environmental exposure data were collected as part of method described by the Department of Housing and Urban their study. Environmental data are particularly useful in Development (HUD, 1990). The microwave oven (CEM providing insights into exposure pathways, and for making MDS-2000) was set at 81% power and the digestion time recommendations on reduction of Pb exposure and thus was 23 min. Pb concentrations were determined using reduction of BPb level. In those Mexican villages where flame atomic absorption (FAA) spectrometer (Perkin-Elmer residents make a living with pottery production, raw Model 3100) at the wavelength of 283.3 nm. The analytical materials used to glaze pottery contain Pb oxide powder detection limit of the FAA was 0.5 ppm. which can be easily spread and are accessible to both adults and children (Avila et al., 1991; Fernandez et al., 1997). Therefore, the village residents may receive Pb exposures Dust Samples The dust samples were obtained using a through many or all of the following major pathways: (a) commonly used method which consisted of wiping areas of inhalation of Pb-containing airborne particles; (b) ingestion known size with commercially available moist towelettes of Pb-contaminated dust and soils; and (c) ingestion of Pb- (Thomas and Spiro, 1994; Orlova et al., 1995). The size of contaminated water and food. In the present study, we have the area was usually estimated in the present study, because therefore collected data on each of these Pb exposure flat continuous areas such as tabletops usually did not exist pathways through a methodology that includes samples in the sampled homes. Typical surfaces sampled were taken from air, dust, soil, and foods in Pb-containing boards for tortilla production, water jugs, shelves, and pottery. This program provides a unique set of insights and utensils. Hands and clothing of people loading the kiln or thus also recommendations on how to reduce Pb exposures glazing were also sampled for dust. According to previous resulting from the pottery production and use. studies using this dust collection method, this method may underestimate the Pb loading (Chavalitnikul and Levin, 1984). After drying in the laboratory, each towelette was Methods placed in a 120-ml digestion liner with 30 ml of 20% nitric acid added and then digested in the CEM microwave. After Study Location and Population the digestion, 30 ml of deionized water were added to the The present study was carried out in the village of Santa Fe sample, and any undissolved towelette residues were de la Laguna (referred to here as `Santa Fe'), MichoacaÂn, removed. Although a tiny amount of the residue was Mexico. The village contains about 950 homes averaging observed, it was felt that the digestion was sufficient to free six persons each. Samples were collected in 17 homes in this the Pb from the wipe matrix. Pb concentrations were village over a 5-week period in July and August of 1996. determined using the same FAA technique as for the soil The 17 homes were selected on the basis of their pottery samples. firing. All homes sampled had soil floors, low walls that allowed passage of air from one room to another, and at most one spigot of running water. The roofs were of varying Air Samples Samples of airborne respirable suspended heights, and usually allowed the smoke to flow into particles (RSP<7.1 m in aerodynamic diameter) were neighbors' homes or into adjoining rooms. The PaÂtzcuaro collected during the glazing and firing of the pottery, using lake region is the site of our ongoing investigation of the an SKC Aircheck pump, cyclone attachment, and 0.5 m environmental and health impacts of biomass and bioenergy matched weight filters. Sampling duration varied and utilization strategies (Saatkampet al., 1998). depended on the duration of the task. Both personal and stationary samples were collected. For personal sampling, the pump was attached to the chest of the person performing Sampling and Analytical Methods various tasks of the pottery production. For stationary sampling, the sampler was placed at the breathing level of a Soil Samples Soil samples were collected at a designated typical adult (1.5±1.7 m above the ground) and in an indoor location inside of a study home. This spot, near the pottery area where the family members would usually gather (1±2 kilns, was in close proximity to the kitchen and cooking m away from the pottery kiln). The sampling flow rate was

344 Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico Hibbert et al.

1869 ml/min in accordance with the manual of the SKC Results and discussion sampler. The filters were then digested using the standard digestion method (HUD, 1990). Pb concentrations were determined using the same FAA technique as for the soil Pb Concentration in Soil and dust samples. The soil Pb concentrations measured in the 17 homes are shown in Table 1 . The concentrations ranged from 0.39 (House nos. 7 and 16) to 19.8 mg/g (House no. 12) with a Pottery Leachate Samples Several pots made in the village median of 3.42 mg/g. were tested for leachable Pb in Environmental and Pb does not biodegrade, decay or break down in soil or Occupational Health Sciences Institute (EOHSI) labora- dust. Even if the primary source is removed, the soil can act tories. The test method was adapted from the draft protocol as a source for hundreds of years. A rise of 1 mg/g Pb from the International Standard Organization (ISO) proce- content in soil or dust causes a 3±7 g/dl BPb increase in dure (ASTM, 1994; ISO, 1996). Each tested pot was filled children, although the uncertainty regarding a general with 4% acetic acid and set for 24 h. However, the tested relationshipbetween soil Pb level and BPb in children is pots started leaking through their walls or tiny cracks only large (EPA, 1986; ATSDR, 1988). It is therefore difficult to after a few hours. The portion of the acid solution coming set Pb concentration standards for soil in the US where soil out of a pot was collected and mixed with the portion intake is generally a more important exposure pathway than remaining inside of the pot. The Pb content in the mixture inhalation for children. However, EPA recommends the was determined with a method using inductively coupled following guidelines: if the concentration is less than 400 plasma-mass spectroscopy (ICP-MS) (Fisons Plasma Quad ppm (i.e., 0.4 mg/g) Pb in the soil, then no action is requited. PQS). The detection limit of the ICP-MS was 0.5 ppb. In an area frequently used by children, interim controls should begin for soils with Pb levels between 400 (0.4 mg/ g) and 5000 ppm (5 mg/g). Interim controls include Cooked Bean Stew Samples In Mexican villages, beans are spreading mulch, shrubs, or gravel over the soil. If soil Pb very common in everyday's diet for both children and concentrations are above 5000 ppm (5 mg/g), abatement adults. In this study, we attempted to examine potential Pb should be undertaken. This would entail permanent removal contamination of foods cooked in the pots made in the and replacement of the top several centimeters of soil, or village. Before cooking, the pots were rinsed once with tap covering with cement paving or brick. Growing grass over water. A mixture was made containing beans, a few drops of corn oil, tap water (half volume of the pot), hot picked peppers, table salt, and one small tomato. From this batch of bean stew, half was put into the Pb-glazed pot and was Table 1. Pb concentration in soil samples. cooked for about 1 h, and half was put in a stainless steel pot and cooked for the same amount of time. After the stews House number Soil Pb (mg/g) cooled off, the stew cooked in the stainless steel pot was 1 3.21 homogenized in a blender; the stew cooked in the Pb-glazed 2 0.53 pot was separated into three portions: one from the top, one 3 1.43 from the middle, and the other from the bottom of the pot. 4 6.34 Each portion was then homogenized. A small fraction (7±13 5 0.99 g) of each homogenized stew was digested with 30 ml of 6 7.63 20% nitric acid in the CEM microwave. After the digestion, 7 0.39 30 ml of deionized water was added to the sample before 8 14.5 being analyzed using the ICP-MS. 9 0.54 10 13.7 11 3.42 Quality Control Both FAA and ICP-MS were calibrated on 12 19.8 a regular basis, according to the standard calibration 13 12.3 schedule and procedure, with standards from National 14 3.42 Institute of Standards and Technology (NIST). The NIST 15 18.6 reference materials 2709 and 2711 were used as quality 16 0.39 assurance checks for all samples. Sample digestion blanks, 17 2.1 reagent blanks, and Pb solution spikes were included in all Mean 6.42 analytical runs. Although acceptable instrument error was S.D. 6.73 set at 20%, most QC analyses were within 10%. Median 3.42

Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) 345 Hibbert et al. High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico the soil can be effective if the grass is maintained and if it (House no. 8), by a factor of 21; that on cooking utensils prevents contact with the soil (EPA, 1995). ranged by a factor of 13; and that on cookstoves ranged by a In Santa Fe Village, two out of 17 homes exhibited Pb factor of 4. concentrations of 0.39 mg/g; 12 houses (71%) had soil Pb Among all the surfaces sampled, the lowest Pb level was levels above 1 mg/g; 11 houses (64%) had soil Pb levels 172 g/f t 2 for cooking utensils of House no. 7. This level is above 2 mg/g; and five houses (29%) had soil Pb levels even even above the EPA-recommended value of 100 g/f t 2 for above 10 mg/g. These soil samples were collected inside of uncarpeted floors (EPA, 1995). (Obviously, utensils would the homes on soil floors where children often play. be associated with a higher exposure even if the floor and Therefore, according to the EPA guidelines, controls and utensils had the same Pb loading.) In response to Pb abatement should be undertaken in all of these homes. The exposures from paint and dust in urban homes, the EPA control techniques recommended by EPA, however, are not does provide a set of guidelines for window sills (500 g/ necessarily applicable to these homes, because the EPA ft2) and window troughs/wells (800 g/f t 2) (EPA, 1995). control strategies are primarily for outdoor soils, not for These standards are also not of direct relevance to indoor soil floors. developing nation settings, as the household items on which we measured Pb concentrations at harmful levels come into human contact more frequently than do the Pb Loading on Household Surfaces window sill, trough, and wells examined by the EPA. In In the 17 sampled homes, flat continuous areas such as addition, high Pb loading was also found on the clothes, tabletops usually did not exist and simple furniture was hands, and tools (glaze rock) of the workers (see Table 2). lacking in some of the homes. Therefore, our dust sampling strategy was to collect surface dust on household items that the residents might most frequently contact with. The Pb Concentration in Air sampled items, along with the Pb loading, are shown in Pb concentrations in respirable suspended particles Table 2. The results indicate that Pb loading had a large (RSP<7.1 m in aerodynamic diameter) were measured variation with respect to different surfaces (items) both using both personal sampling and stationary sampling within a house and across different houses. For example, Pb methods. Personal samples were collected for eight workers in surface dust ranged from 690 (on cookstove) to 33 060 (all men) while performing the pottery firing and for 15 g/f t 2 (on shelves), by a factor of almost 50, within House workers (mainly women) while performing the pottery no. 3. Across different homes, the surface dust Pb on tables/ glazing. Personal Pb concentrations for the eight firing chairs ranged from 264 (House no. 4) to 5530 g/f t 2 workers ranged from 10.7 (House no. 2) to 234 g/m3

Table 2. Dust Pb loading (g/f t 2) measured in 17 homes.

House Shelves Worker's Worker's Glaze Table/ Wall Utensils Water Mugs Bread Tortilla Stove Average number hands cloth rock Chair jugs oven maker 1 2262 2262 2 12 948 2280 414 1020 3975 4127 3 33 060 8754 1260 12 216 1038 690 9503 4 264 600 1440 546 1920 954 5 6 46 260 870 2256 16 462 7 172 172 8 12 012 5530 396 2344 5071 9 540 540 10 11 1050 1050 12 13 1800 19 020 2940 7920 14 24 120 24 120 15 8826 708 780 2940 3314 16 17 1674 5040 912 3074 2675 A blank cell indicates that no sample was collected.

346 Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico Hibbert et al.

(House no. 5) with a median of 27.3 g/m3. Personal Pb Table 4. Results from Pb leachate tests. concentrations for the 15 glazing workers ranged from the undetectable (Houses nos. 10 and 12) to 454 g/m3 (House Pot sample Volume of Pb concentration no. 11) (see Table 3). The large variation in these personal number pot (ml) (mg/l) Pb concentrations was perhaps mainly due to the difference A 883 6.99 in glaze ingredients among different household manufac- B 250 158 tories and behavioral differences among different workers. C 262 39.1 Results from the stationary (indoor) samples indicate that D 910 29.0 airborne Pb concentrations were lower during nonglazing E 931 12.4 period compared to glazing period (see Table 3). During the pottery glazing, airborne Pb concentrations in 12 houses measured ranged from 15.3 (House no. 5) to 104 g/m3 mugs, and cooling wares, posting a significant risk for Pb (House no. 13) with a median of 56.1 g/m3. contamination of water and foods stored or cooked in such pottery.

Pb Leachate from the Pottery Pb Concentrations in Cooked Bean Stews Residents of Santa Fe consume jug-stored water, never tap A previous study demonstrated that people who consume water. Thus, there should not be any concerns on Pb stew made in Pb-glazed pottery are more likely to have high contribution from plumbing. The water containers, how- BPb levels of >15 g/dl (Avila et al., 1991). In Santa Fe, like ever, are normally Pb-glazed jugs manufactured locally. many parts of Mexico, bean stews are very common in the These containers may cause Pb leaching to the water stored. everyday diets for both children and adults. Therefore, we To test potential Pb leaching from the pottery made in Santa conducted tests of Pb content in bean stews cooked in a pot Fe, we have used the standard pottery testing method made in the village to provide representative information on established by the ISO (ASTM, 1994; ISO, 1996). Five pots potential Pb contamination from cooking practices. The test were tested for Pb leaching. The results are shown in Table results (shown in Table 5) indicate not only that the bean 4, indicating that the Pb leachate concentrations ranged stew cooked in the Pb-glazed pot had a Pb concentration widely (from 7 to 158 mg/l) among the five pots tested. about 25 times higher than that cooked in the stainless steel These Pb leachate levels far exceeded ISO permissible pot. In addition, the food portion obtained from the bottom limits of 0.5±2.0 mg/l for Pb in the hollow ware, cups, of the pot exhibited an even higher Pb concentration than those obtained from the middle and topof the pot. It is important to note that the bean stew ingredients also Table 3. Pb concentration in air (g/m3). made a contribution to total Pb in cooked stew, because Pb content in the bean stew cooked in the stainless steel pot was House Personal Stationary Distance from higher than the Pb concentration in the tapwater. Indeed, Pb number During During Glazing No kiln (m) has been detected in a variety of foods in the US. Typical firing glazing glazing concentrations of Pb in various foods ranged from 0.002 to 1 155 51.7 16.9 1.6 0.649 g/g (DHHS, 1998). Pb concentration in the stew 2 10.7 35.7 ± 3 35.7 293 53.5 ND 1.5 4 15.5 47.6 N/A Table 5. Pb concentration in Bean Stews Cooked in a Pb-glazed and a 5 234 44.6 15.3 N/A stainless steel pot. 6 17.8 17.8 19.3 N/A 7 29.2 13.4 Sample Description of samples Pb concentration 8 61.1 34.1 1.2 number (g/g) 9 17.3 13.1 W1 Tapwater from EOHSI laboratory 0.0012 10 25.3 ND W2 Tapwater from EOHSI laboratory 0.0013 11 454 90.1 25.2 1.1 L2 Cooked bean stews from the bottom of the 3.8 12 ND 58.6 ND 1.0 Pb-glazed pot 13 104 1.2 L3 Cooked bean stews from the topof the 2.0 14 26.1 17.4 2.0 Pb-glazed pot 15 28.2 61.0 ND 2.0 L4 Cooked bean stews from the middle of the 1.4 16 32.7 ND 1.0 Pb-glazed pot 17 14.9 58.8 13.1 1.3 S1 Cooked bean stews form a stainless steel pot 0.12 A blank cell indicates that no sample was collected; ND, not detected. S2 Cooked bean stews form a stainless steel pot 0.13

Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) 347 Hibbert et al. High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico cooked in the stainless steel pot fell into this typical range. of 20 m3/day or 0.83 m3/h (AIHC, 1994; EPA, 1997), we Pb concentration in the stew cooked in the Pb-glazed pot, infer a daily Pb exposure via inhalation during the 23.14-h however, exceeded the high end of this range. This suggests nonfiring/glazing period of: that use of Pb-glazed pottery of the tested type as cooking vessel could cause significant Pb contamination of cooked 0:5 g=m3Â0:83 m3=hÂ23:14 h 9:6g: foods. The Pb leaching rate might be the greatest when the Therefore, the daily Pb exposure via inhalation for the pottery is new and decreases as the pottery is getting used. firing worker and glazing worker would be 102 and 106 g, Even so, given that this type of pottery usually has a short respectively, or 104 g, on average, for the Santa Fe adult `lifetime' and that used ones may have cracks and thus may residents. release more Pb from the glaze to foods, Pb contamination Small children living in the Santa Fe households potential is expected to be high regardless of new or used normally remain in their family or a neighbor's household pottery. most of the time. Childhood daily Pb exposure via inhalation can be estimated as follows: median respiration rate=17.3 m3/day or 0.72 m3/h (AIHC, 1994); 0.86 h Pb Exposures via Three Major Pathways 3 Human exposure to environmental pollutants depends not exposure to Pb at 56.1 g/m (median indoor levels during the firing/glazing). Combining this with the 23.14-h only on concentrations of the pollutants, but also on contact 3 rates which are determined by time/activity patterns (Lioy, exposure duration to Pb at 0.5 g/m (median indoor levels 1990; Lanphear and Roghmann, 1997). In Santa Fe, Pb was during nonfiring/glazing) translates to a daily Pb exposure found in multiple environmental media (soil, dust, air, and of 43 g/person. prepared foods) which could result in Pb exposures via multiple pathways (soil/dust ingestion, inhalation, food Exposure via Soil Ingestion Although many regulatory- ingestion). To estimate multimedium Pb exposures for based exposure assessments use a soil ingestion rate of 200 adults and children of Santa Fe, we use data collected in this mg/day for children and 100 mg/day for adults, the soil study and adapt some necessary data published in the ingestion rate should vary largely for different individuals literature, mainly in the Exposure Factors Handbook (EPA, living under different conditions or with different personal 1997) and Exposure Factors Sourcebook (AIHC, 1994). habits. For example, the median soil ingestion levels reported by different investigators range from 11 to 618 Exposure via Inhalation Our survey indicates that a Santa mg/day for US children and from 1 to 77 mg/day for adults Fe household engaged in pottery production typically fires (Binder et al., 1986; Clausing et al., 1987; Calabrese and the kiln once or twice per week, and that the average firing/ Stanek, 1991; Calabrese et al., 1989a, 1989b; Thompson glazing time is approximately 6 h/week or 0.86 h/day. From and Burmaster, 1991). These values were based upon there, we can estimate total worker exposure. studies of US adults and children. If the respiration (inhalation) rate for an adult performing The soil/dust ingestion rates for children are expected to strenuous activity (which is appropriate for workers be very different in Santa Fe from those reported for US performing the firing/glazing of the pottery) is assumed to populations. The existence of soil floors means that children be 3.9 m3/h (AIHC, 1994), a worker's Pb exposure during in the village households have access to dirt no matter what the firing process would be: the weather or the time of day. Hand±mouth and soil ingestion behaviors also probably differ greatly between the US and Mexico. The homes having dirty floors as the only 27:3 g=m3 medianÂ3:9m3=hÂ0:86 h 92g; place children can be set down is one thing that contributes to this. In addition, due to lack of running water, hygiene is and a glazing worker's Pb exposure during the glazing very different, and children are often constantly dirty. Also, would be: children do not have cribs, playpens or other devices to contain them. Once they are mobile, children have access to 28:2 g=m3 medianÂ3:9m3=hÂ0:86 h 95g: any area in the house. Cultural differences exist that make a typical mother in Santa Fe less concerned about her children playing in dirty areas, or putting dirty objects or soil into Measured Pb concentrations during nonfiring/glazing their mouths. hours ranged from ND to 34.1 g/m3 (see Table 3) with a Differences in hygiene also contribute to variations median of ND. We assume ND to be half of the method between soil and dust ingestion rates for adults in Santa Fe detection limit of approximately 1 g/m3 and get a median as compared to those expected in the US. The lack of indoor air Pb concentration for nonfiring/glazing period of 0.5 g/ plumbing encourages people to bathe less frequently, and m3. Using the median point value for adult respiration rate not to wash their hands as frequently. The fact that very few

348 Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico Hibbert et al. adults in Santa Fe knew that the pottery materials constantly more prevalent than now (EPA, 1994). Our rough estimate in contact with their hands were toxic, would also mean that did not consider Pb exposure through stored drinking. they would take less precautions to avoid ingestion than an Based on the above estimation, total daily Pb exposure average person in the US. In addition, dirty clothes and hair (intake) is 4.0 mg for adults and 3.4 mg for children living in can serve as a constant source to hands, even if the hands Santa Fe. The relative contribution of each exposure were washed regularly. In sum, our calculations of child- pathway to the total exposure is shown in Figure 1. The hood and adult Pb ingestion via environmental exposure are exposure values estimated in this paper are potential doses, likely to be quite conservative. without reference to bioavailability (body absorption rate) For a `typical' child in Santa Fe ingesting 293 mg of soil/ between inhaled Pb and ingested Pb (Renner, 1995; Adgate, dust, the daily Pb exposure would be 293 mgÂ3.42 mg/g=1 1996). We realize that the exposure estimates made in this mg. Assuming an adult in Santa Fe ingested 77 mg of soil/ paper are relatively rough due to the small sample size and dust, the daily Pb exposure would be 77 mgÂ3.42 mg/ large uncertainties associated with data on the time/activity g=0.3 mg. In this estimate, we took the soil Pb to be the patterns and exposure contact rates. We therefore stress that median for the 17 houses sampled. Surface dust Pb-loading while suggestive of serious health concerns, any quantita- values (mg/ft2) were not used in the estimation because: (a) tive conclusion and recommendation drawn from this paper it is not possible to know the fraction of dust or soil in the should be considered preliminary. More comprehensive total amount of soil/dust ingested; (b) dust Pb concentra- studies linking blood Pb levels and the sorts of environ- tions (mg/g) were not measured directly; and (c) the mental exposure work reported here are warranted. difference between the soil Pb concentration and the surface Pb concentration should be relatively small, given that the soil samples were taken from the surface layer of the dirty Conclusions and recommendations floors. It was estimated that only in the Purhepecha Region, there Exposure via Food Ingestion Total dietary intake of an are almost 3900 small pottery enterprises (Masera et al., average adult, based on surveys conducted in the US and 1998) and that in the entire Mexico, there are approximately Europe, ranges from 1.4 kg/day for the UK to 1.6 kg/day for >1.5 million glaze potters (FONART, 1997). Due to the use the US (AIHC, 1994). Substantially different values may be of Pb oxide in pottery production, extremely high Pb apparent worldwide, depending on food availability and diet concentrations were measured in personal and indoor air composition. For the Santa Fe residents, we assume that the samples, household surface dust samples, and household food consumption rate was 1.5 kg/day for adults and 1 kg/ soil samples in Santa Fe Village. A very high Pb content was day for children. Using average Pb content of the cooked found in the bean stew cooked in a pot made in the village, bean stew (2.4 g/g), we estimate that the daily Pb exposure leading to a high estimated Pb intake from consumption of via food ingestion would be 3.6 mg for adults and 2.4 mg for foods for Santa Fe residents and any user of this type of children. This estimated Pb intake for Santa Fe children is cookware. These values indicate a very significant public more than an order of magnitude higher than the estimate of health problem for the village residents and other similar Pb intake from consumption of local produce by US communities. For example, our brief survey found that the children of 2±6 years in 1983 when Pb was substantially following health complaints were common in the village: discomfort in hands, joints, and backs, especially kidney region of back, underweight children and adults, respiratory infections, parasites, diarrhea, and seizures in children. Some of these ailments could potentially be related directly or indirectly to Pb exposure. In the village, most homes serve both as dwellings and as pottery production facilities. Each aspect of the process, including making the glaze, glazing the pieces, and firing in rustic, open kilns, is performed in the presence of, and sometimes with the helpof, children and often occurs within several meters of the food preparation and eating areas. Therefore, segregating work areas with residential areas would reduce Pb exposure via inhalation and soil/dust ingestion. Because the contribution from soil/dust ingestion is larger than from inhalation, especially for children (see Figure 1. Estimated daily Pb exposures (mg/day) via three major Figure 1), it would be important to prevent the Pb- exposure pathways for the Santa Fe residents. containing glaze powder from spreading. None of personal

Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) 349 Hibbert et al. High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico protection has been employed during glazing or firing in the 1996. Rutgers, the State University of New Jersey and Robert Wood Santa Fe workers. Some simple personal protection such as Johnson Medical School, Piscataway, NJ, USA, 1996. AIHC. Exposure Factors Sourcebook. American Industrial Health Council, wearing masks, however, can be expected to reduce the Washington, DC, USA, 1994. inhalation exposure. Some behavioral habits of the village ASTM. ASTM Test Method for Lead and Cadmium Extraction from residents, e.g., are expected to enhance the exposure. For Glazed Ceramic Surfaces, C738-94. American Society for Testing and example, the presence of children and babies near the Pb Materials, Philadelphia, PA, USA, 1994. glaze results in children being tended to by parents with Pb- ATSDR. The Nature and Extent of Lead Poisoning in Children in the United States: A Report to Congress. Agency for Toxic Substances and covered hands. The pottery kilns serve as a source of heat Disease Registry, Atlanta, GA, USA, 1988. and light, and therefore a place to gather socially or to eat Avila M.H., Romoeu I., Rios C., Rivero A., and Palazuelos E. Lead-glazed meals. Lack of running water and thus lack of regular ceramics as major determinants of blood lead levels in Mexican hygiene increase the exposure. Changes in these behavioral women. Environmental Health Perspectives 1991: 94: 117±120. habits would lead to a reduction of their Pb exposure (Lioy Binder S., Sokal D., and Maughan D. Estimating the amount of soil et al., 1998). ingested by young children through tracer elements. Archives of Environmental Health 1986: 41: 341±345. Surprisingly, our rough estimates suggest that Santa Fe Calabrese E.J., and Stanek III E.J. A guide to interpreting soil ingestion residents may ingest more Pb from foods cooked in the Pb- studies: II. Qualitative and quantitative evidence of soil ingestion. glazed pottery than from their household soils and dust. This Regulatory Toxicology and Pharmacology 1991: 13: 278±292. means, unfortunately, that not only the Santa Fe residents, Calabrese E.J., Barnes R., Stanek III E.J., Pastides H., Gilbert C.E., but also anyone who uses the pottery as cookware, will be Veneman P., Wang X., Lasztity A., and Kostecki P.T. How much soil potentially at a high risk of Pb exposure. To eliminate or do young children ingest: an epidemiological study. Regulatory Toxicology and Pharmacology 1989a: 10: 123±137. reduce such a risk, thus, the feasibility of alternative glazes Calabrese E.J., Pastides H., Barnes R., Edward C., Kostecki P.T., Stanek needs to be explored. III E.J., Veneman P., and Gilbert C.E. How much soil do young Among all the recommendations, however, the first and children ingest: an epidemiologic study. In: Calabrese E.J., Kostecki most important step is perhaps the communication of the P.T. (Eds.), Petroleum-Contaminated Soils, Vol. 2, Chap. 30, 1989b, risk to those pottery workers and the collaboration with pp. 363±397. them to best modify or even change the production process. CDC. Preventing Lead Poisoning in Young Children: A Statement by the Centers for Disease Control. Centers for Disease Control and The communication and public education have been Prevention, Atlanta, GA, USA, 1991. recognized to be the key of preventing childhood Pb- Chavalitnikul C., and Levin L. A laboratory evaluation of wipe testing poisoning in the US urban populations (Goldman, 1997). based on lead oxide surface contamination. Am. Ind. Hyg. Assoc. J. Without recognition of their health problems caused by Pb 1984: 45 (5): 311±317. exposure and help from an outside source, these village Clausing P., Brunekreef B., and Van Wijnen J.H. A method for estimating soil ingestion by children. International Archives of Occupational and residents are too caught upin the daily task of carrying out Environmental Health 1987: 59: 73±82. their work to provide what little they can for their families to DHHS. Toxicological profile for lead. US Department of Health and find a new means of income. A way should be found for Human Services, Washington, DC, USA, 1998, pp. 334±339. them to keeptheir traditional occupation in a safe way, or Diaz-Barriga F., Batres L., and Calderon J. The El Paso smelter 20 years provide training or assistance for other occupations. later: residual impact on Mexican children. Environmental Research 1997: 74: 11±16. EPA. Air Quality Criteria for Lead, Vol. I±IV. EPA-600/8-83-028F. Acknowledgments Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment, US Environmental Protection Agency, The authors thank Dr. John Adgate, Dr. Rufus Edwards, and Research Triangle Park, NC, USA, 1986. Dr. Tom Wainman for their comments and Dr. Lih-Ming EPA. Guidance Manual for the Integrated Exposure Uptake Biokinetic Yiih for his assistance in the sample analysis. Thanks also to (IEUBK) Model for Lead in Children. EPA-540-R-93-081. Office of Jose Luis Fabien and the residents of Santa Fe for their Research and Development, US Environmental Protection Agency, Research Triangle Park, NC, USA, 1994. invaluable helpin the field study portionof this research. EPA. Guidance on Identification of Lead-Based Paint Hazards: Notice, This study is supported by the MacGregor Fund and the Sept. 11, 1995. Office of Research and Development, US Environ- Alexander Polsky Fund at Princeton University. Dr. J. mental Protection Agency, Washington, DC, USA, 1995. Zhang's study was also supported by NIEHS Center Grant EPA. Exposure Factors Handbook. EPA/600/P-95/002Fa-c. Office of no. 05022. Dr. D.M. Kammen's work was supported by the Research and Development, US Environmental Protection Agency, Washington, DC, USA, 1997. Class of 1934 Preceptorship and grants from the Summit Fernandez G.O., Martinez R.R., Fortoul T.I., and Ralazuelos E. High Compton and Foundations. blood lead levels in Ceramic Folk Art Workers in Michoacan, Mexico. Archives of Environmental Health 1997: 52: 51±55. FONART, AboletõÁn de Prensa 19 de febrero (1997). In: UKATA (Revista References del Arte Popular Michoacano), No. 13, AOÁ o III, Marzo ± Abril. Morelia, Mexico, 1997. Adgate J. House dust lead: relationships between exposure metrics and Goldman L.R. Information: the key to preventing childhood lead determination of source contributors. Ph.D. Dissertation, October, poisoning. Journal of Environmental Health 1997: 59: 45±46.

350 Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) High lead exposures resulting from pottery production in a village in MichoacaÂn State, Mexico Hibbert et al.

Harrison R.M., and Laxen D.P.H. Lead Pollution: Causes and Control. Murgueytio A.M., Evans R.G., and Sterling D. Behaviors and blood lead Cambridge Univ. Press, London, UK, 1981. levels of children in a lead-mining area and a comparison community. HUD. Lead-Based Paint: Interim Guidelines for Hazard Identification and Journal of Environmental Health 1998: 60: 14±20. Abatement in Public and Indian Housing. Department of Housing and Orlova A.O., Banon D.I., Farfel M.R., Thomas V.M., Aleschukin L.V., Urban Development Washington, DC, USA, 1990. Kudashov V.V., Shine J.P., and Kruchkov G.I. Pilot study of sources of SO. Draft International Standard for Ceramic Ware, Glass±Ceramic Ware, lead exposure in Moscow, Russia. Environmental Geochemistry and and Glass Flatware in Contact With FoodÐRelease of Lead and Health 1995: 17: 200±210. Cadmium, Part 1: Method and Part 2: Permissible Limits. ISO/TC 166, Pirkle J.L., Brody D.J., Gunter E.W., Kramer R.A., Paschal D.C., Flegal International Standard Organization, Secretariat ANSI, 1996. K.M., and Matte T.D. The decline of blood lead levels in the United Lanphear B.P., and Roghmann K.J. Pathways of lead exposure in urban States: the National Health and Nutrition Examination Survey children. Environmental Research 1997: 74: 67±73. (NHANES). J. Am. Med. Assoc. 1994: 272: 284±291. Lioy P.J. Assessing total human exposures to contaminants. Environmental Renner R. When is lead a health risk? Environmental Science and Science and Technology 1990: 24: 938±945. Technology 1995: 29: 256A±261A. Lioy P.J., Yiin L., Adgate J., Weisel C., and Rhoads G.G. The effectiveness SaatkampB.D., Masera O., and Kammen D.M. Social versus technical of a home cleaning intervention strategy in reducing potential dust and visions of the energy ladder: fuels, stoves, and indoor air pollution in lead exposures. Journal of Exposure Analysis and Environmental JaraÂcuaro, MeÂxico. Boiling Point 1998: 40: 16±18. Epidemiology 1998: 8: 17±35. Thomas V., and Spiro T. Emissions and exposure to metals: cadmium and Masera O., Masera D., and Navia J. DinaÂmica y Uso de los Recursos lead. In: Socolow, et al. (Eds.), Industrial Ecology and Global Change. Forestales en la RegiuÃn Purepecha: El Papel de las PequeoÁas Empresas Cambridge Univ. Press, Cambridge, 1994, pp. 297±315. Artesanales. Documento de Trabajo No. 22, GIRA, AC PaÂtzcuaro, Thompson K.M., and Burmaster D.E. Parametric distributions for soil Mexico, 1998. ingestion by children. Risk Analysis 1991: 11: 339±342.

Journal of Exposure Analysis and Environmental Epidemiology (1999) 9(4) 351